RESUMO
AIMS: The aims of the study were to identify predictors of locoregional failure (LRF) following surgery for pancreatic adenocarcinoma, develop a prediction risk score model of LRF and evaluate the impact of postoperative radiation therapy (PORT) on LRF. MATERIALS AND METHODS: A retrospective review was conducted on patients with stages I-III pancreatic adenocarcinoma who underwent surgery at our institution (2005-2016). Univariable and then multivariable analyses were used to evaluate clinicopathological factors associated with LRF for patients who did not receive PORT. The risk score of LRF was calculated based on the sum of coefficients of the predictors of LRF. The model was applied to the entire cohort to evaluate the impact of PORT on the high- and low-risk groups for LRF. RESULTS: In total, 467 patients were identified (median follow-up 22 months). Among patients who did not receive PORT (n = 440), predictors of LRF were pN+, involved or close ≤1 mm margin(s), moderately and poorly differentiated tumour grade and lymphovascular invasion. After adding patients who received PORT, the 2-year LRF in the high-risk group was 57% for patients who did not receive PORT (n = 242) and 32% among patients who received PORT (n = 22), with an absolute benefit to LRF of 25% (95% confidence interval 5-52%, P = 0.07). The 2-year overall survival for the high-versus the low-risk group was 36% versus 67% (P < 0.001). CONCLUSION: This risk group classification could be used to identify pancreatic adenocarcinoma patients with higher risk of LRF who may benefit from PORT. However, validation and prospective evaluation are warranted.